Symmetry, Saddle Points, and Global Optimization Landscape of Nonconvex Matrix Factorization
From MaRDI portal
Publication:5224010
DOI10.1109/TIT.2019.2898663zbMath1432.90123arXiv1612.09296OpenAlexW2931810883WikidataQ128135569 ScholiaQ128135569MaRDI QIDQ5224010
Han Liu, Zhaoran Wang, Tuo Zhao, Raman Arora, Junwei Lu, Jarvis D. Haupt, Xing Guo Li
Publication date: 19 July 2019
Published in: IEEE Transactions on Information Theory (Search for Journal in Brave)
Full work available at URL: https://arxiv.org/abs/1612.09296
phase retrievallow-rank matrix factorizationdeep linear neural networksclass of machine learning problemslandscape of nonconvex optimization with underlying symmetric structureslocations of stationary pointsmatrix sensing problemnull space of Hessian matrices of objective function
Related Items
Model-free Nonconvex Matrix Completion: Local Minima Analysis and Applications in Memory-efficient Kernel PCA ⋮ Role of sparsity and structure in the optimization landscape of non-convex matrix sensing ⋮ Column $\ell_{2,0}$-Norm Regularized Factorization Model of Low-Rank Matrix Recovery and Its Computation ⋮ GNMR: A Provable One-Line Algorithm for Low Rank Matrix Recovery ⋮ Analysis of the optimization landscape of Linear Quadratic Gaussian (LQG) control ⋮ Implicit regularization in nonconvex statistical estimation: gradient descent converges linearly for phase retrieval, matrix completion, and blind deconvolution ⋮ Median-Truncated Gradient Descent: A Robust and Scalable Nonconvex Approach for Signal Estimation ⋮ The global optimization geometry of shallow linear neural networks ⋮ Nonconvex Robust Low-Rank Matrix Recovery ⋮ Error bound of critical points and KL property of exponent 1/2 for squared F-norm regularized factorization ⋮ Provable accelerated gradient method for nonconvex low rank optimization ⋮ $$L^p$$ Continuity and Microlocal Properties for Pseudodifferential Operators ⋮ Finding Low-Rank Solutions via Nonconvex Matrix Factorization, Efficiently and Provably ⋮ Unnamed Item